Recently, portable terminals, such as portable telephone, in which emphasis is placed on portability, are in widespread use, such that, for improving portability, reduction in size and weight of the portable terminal has become a desideratum. In order to improve the portability in the portable terminal, an input device, such as a keyboard, has to be reduced in size or eliminated. So, in the portable terminal, a speech recognition device, which is less bulky than the conventional input device, is stirring up notice. That is, with the use of such speech recognition device, not only can the terminal itself be reduced in size, but also the terminal can be used by a user with improved facility.
On the other hand, in the contents selection system, constructed between the server and the client, development of the digital signal processing technique in the technique of compressing data, such as speech, or in the field of broadcasting/communication, has led to realization of the service of furnishing the audio information, such as the so-called music-on-demand, referred to below as MOD.
However, if a speech recognition device is used in the portable terminal as means for recognizing the input information, it is extremely difficult to eliminate errors in speech recognition. Moreover, the probability of accurate recognition of the input audio information in the conventional speech recognition device is not that high.
Moreover, if, in the MOD service, the speech recognition device has made mistaken recognition of the speech information input for designating the music desired to be furnished, the user has to input the same speech information to the speech recognition device. In addition, since the speech recognition device cannot necessarily make correct recognition of the speech the user inputs next, the user has to input the same speech information repeatedly to the speech recognition device, thus seriously inconveniencing the user.
Moreover, if the user has specified the music etc desired to be furnished by inputting several pieces of the speech information to the speech recognition device, it may be an occurrence that plural music numbers etc are meeting with the conditions of the input speech information. In such case, it is extremely difficult to specify the music which the user desires to be furnished.